Commit
•
63c8c98
1
Parent(s):
76c9b47
model uploaded
Browse files- config.json +51 -0
- merges.txt +0 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scaler.pt +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +1 -0
- tokenizer.json +0 -0
- tokenizer_config.json +1 -0
- trainer_state.json +1626 -0
- training_args.bin +3 -0
- vocab.json +0 -0
config.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "facebook/bart-large-mnli",
|
| 3 |
+
"_num_labels": 3,
|
| 4 |
+
"activation_dropout": 0.0,
|
| 5 |
+
"activation_function": "gelu",
|
| 6 |
+
"add_final_layer_norm": false,
|
| 7 |
+
"architectures": [
|
| 8 |
+
"BartForConditionalGeneration"
|
| 9 |
+
],
|
| 10 |
+
"attention_dropout": 0.0,
|
| 11 |
+
"bos_token_id": 0,
|
| 12 |
+
"classif_dropout": 0.0,
|
| 13 |
+
"classifier_dropout": 0.0,
|
| 14 |
+
"d_model": 1024,
|
| 15 |
+
"decoder_attention_heads": 16,
|
| 16 |
+
"decoder_ffn_dim": 4096,
|
| 17 |
+
"decoder_layerdrop": 0.0,
|
| 18 |
+
"decoder_layers": 12,
|
| 19 |
+
"decoder_start_token_id": 2,
|
| 20 |
+
"dropout": 0.1,
|
| 21 |
+
"encoder_attention_heads": 16,
|
| 22 |
+
"encoder_ffn_dim": 4096,
|
| 23 |
+
"encoder_layerdrop": 0.0,
|
| 24 |
+
"encoder_layers": 12,
|
| 25 |
+
"eos_token_id": 2,
|
| 26 |
+
"forced_eos_token_id": 2,
|
| 27 |
+
"gradient_checkpointing": false,
|
| 28 |
+
"id2label": {
|
| 29 |
+
"0": "contradiction",
|
| 30 |
+
"1": "neutral",
|
| 31 |
+
"2": "entailment"
|
| 32 |
+
},
|
| 33 |
+
"init_std": 0.02,
|
| 34 |
+
"is_encoder_decoder": true,
|
| 35 |
+
"label2id": {
|
| 36 |
+
"contradiction": 0,
|
| 37 |
+
"entailment": 2,
|
| 38 |
+
"neutral": 1
|
| 39 |
+
},
|
| 40 |
+
"max_position_embeddings": 1024,
|
| 41 |
+
"model_type": "bart",
|
| 42 |
+
"normalize_before": false,
|
| 43 |
+
"num_hidden_layers": 12,
|
| 44 |
+
"output_past": false,
|
| 45 |
+
"pad_token_id": 1,
|
| 46 |
+
"scale_embedding": false,
|
| 47 |
+
"torch_dtype": "float32",
|
| 48 |
+
"transformers_version": "4.15.0",
|
| 49 |
+
"use_cache": true,
|
| 50 |
+
"vocab_size": 50265
|
| 51 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:916d118f86a3d9ff866b13b62785e3a9f3b0f2033543c93551e77438ba24d6a3
|
| 3 |
+
size 3250652765
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d141637e022fd2a3fc473dcfd78058485c336bb7c6b67ba249d8673b6beebf51
|
| 3 |
+
size 1625569391
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:99e8d4e69c946bc184ca0fc2c27f35773e7ccf6f0e989a01f37268316a13ec05
|
| 3 |
+
size 16724
|
scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b5561e919eb9db11cdced95c1a49ebda1380b9891404d9cf7a71f688faddb4e
|
| 3 |
+
size 559
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:565c5edf74526a7ab8e0a7cc948a89f3145ac7f2f1781f595a86b47004929861
|
| 3 |
+
size 623
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "sep_token": "</s>", "pad_token": "<pad>", "cls_token": "<s>", "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": false}}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"unk_token": "<unk>", "bos_token": "<s>", "eos_token": "</s>", "add_prefix_space": false, "errors": "replace", "sep_token": "</s>", "cls_token": "<s>", "pad_token": "<pad>", "mask_token": "<mask>", "trim_offsets": true, "model_max_length": 1024, "special_tokens_map_file": null, "name_or_path": "facebook/bart-large-mnli", "tokenizer_class": "BartTokenizer"}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1626 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.807554874936192,
|
| 5 |
+
"global_step": 5500,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.05,
|
| 12 |
+
"eval_gen_len": 11.1391,
|
| 13 |
+
"eval_loss": 1.4340903759002686,
|
| 14 |
+
"eval_rouge1": 23.6512,
|
| 15 |
+
"eval_rouge2": 14.437,
|
| 16 |
+
"eval_rougeL": 21.5776,
|
| 17 |
+
"eval_rougeLsum": 21.9318,
|
| 18 |
+
"eval_runtime": 142.6631,
|
| 19 |
+
"eval_samples_per_second": 35.286,
|
| 20 |
+
"eval_steps_per_second": 1.472,
|
| 21 |
+
"step": 100
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.1,
|
| 25 |
+
"eval_gen_len": 9.006,
|
| 26 |
+
"eval_loss": 1.498289704322815,
|
| 27 |
+
"eval_rouge1": 15.3475,
|
| 28 |
+
"eval_rouge2": 9.3183,
|
| 29 |
+
"eval_rougeL": 14.1923,
|
| 30 |
+
"eval_rougeLsum": 14.233,
|
| 31 |
+
"eval_runtime": 142.2039,
|
| 32 |
+
"eval_samples_per_second": 35.4,
|
| 33 |
+
"eval_steps_per_second": 1.477,
|
| 34 |
+
"step": 200
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 0.15,
|
| 38 |
+
"eval_gen_len": 5.8882,
|
| 39 |
+
"eval_loss": 1.4406640529632568,
|
| 40 |
+
"eval_rouge1": 4.4607,
|
| 41 |
+
"eval_rouge2": 2.5402,
|
| 42 |
+
"eval_rougeL": 4.0061,
|
| 43 |
+
"eval_rougeLsum": 4.0532,
|
| 44 |
+
"eval_runtime": 139.4058,
|
| 45 |
+
"eval_samples_per_second": 36.11,
|
| 46 |
+
"eval_steps_per_second": 1.506,
|
| 47 |
+
"step": 300
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 0.2,
|
| 51 |
+
"eval_gen_len": 8.2084,
|
| 52 |
+
"eval_loss": 1.4275349378585815,
|
| 53 |
+
"eval_rouge1": 11.9263,
|
| 54 |
+
"eval_rouge2": 6.9699,
|
| 55 |
+
"eval_rougeL": 11.0646,
|
| 56 |
+
"eval_rougeLsum": 11.1028,
|
| 57 |
+
"eval_runtime": 141.9092,
|
| 58 |
+
"eval_samples_per_second": 35.473,
|
| 59 |
+
"eval_steps_per_second": 1.48,
|
| 60 |
+
"step": 400
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.26,
|
| 64 |
+
"learning_rate": 1.830866088140208e-05,
|
| 65 |
+
"loss": 1.3664,
|
| 66 |
+
"step": 500
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.26,
|
| 70 |
+
"eval_gen_len": 8.6985,
|
| 71 |
+
"eval_loss": 1.3808070421218872,
|
| 72 |
+
"eval_rouge1": 10.8243,
|
| 73 |
+
"eval_rouge2": 6.646,
|
| 74 |
+
"eval_rougeL": 10.1357,
|
| 75 |
+
"eval_rougeLsum": 10.1831,
|
| 76 |
+
"eval_runtime": 141.9295,
|
| 77 |
+
"eval_samples_per_second": 35.468,
|
| 78 |
+
"eval_steps_per_second": 1.48,
|
| 79 |
+
"step": 500
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.27,
|
| 83 |
+
"learning_rate": 1.824059894503999e-05,
|
| 84 |
+
"loss": 1.2606,
|
| 85 |
+
"step": 520
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 0.28,
|
| 89 |
+
"learning_rate": 1.8172537008677897e-05,
|
| 90 |
+
"loss": 1.2438,
|
| 91 |
+
"step": 540
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.29,
|
| 95 |
+
"learning_rate": 1.8104475072315808e-05,
|
| 96 |
+
"loss": 1.2952,
|
| 97 |
+
"step": 560
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.3,
|
| 101 |
+
"learning_rate": 1.803641313595372e-05,
|
| 102 |
+
"loss": 1.2639,
|
| 103 |
+
"step": 580
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 0.31,
|
| 107 |
+
"learning_rate": 1.796835119959163e-05,
|
| 108 |
+
"loss": 1.2335,
|
| 109 |
+
"step": 600
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 0.32,
|
| 113 |
+
"learning_rate": 1.790028926322954e-05,
|
| 114 |
+
"loss": 1.258,
|
| 115 |
+
"step": 620
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.33,
|
| 119 |
+
"learning_rate": 1.7832227326867452e-05,
|
| 120 |
+
"loss": 1.2501,
|
| 121 |
+
"step": 640
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.34,
|
| 125 |
+
"learning_rate": 1.7764165390505363e-05,
|
| 126 |
+
"loss": 1.2657,
|
| 127 |
+
"step": 660
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 0.35,
|
| 131 |
+
"learning_rate": 1.769610345414327e-05,
|
| 132 |
+
"loss": 1.2659,
|
| 133 |
+
"step": 680
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 0.36,
|
| 137 |
+
"learning_rate": 1.762804151778118e-05,
|
| 138 |
+
"loss": 1.2469,
|
| 139 |
+
"step": 700
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 0.37,
|
| 143 |
+
"learning_rate": 1.7559979581419092e-05,
|
| 144 |
+
"loss": 1.2932,
|
| 145 |
+
"step": 720
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.38,
|
| 149 |
+
"learning_rate": 1.7491917645057003e-05,
|
| 150 |
+
"loss": 1.2521,
|
| 151 |
+
"step": 740
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 0.39,
|
| 155 |
+
"learning_rate": 1.7423855708694914e-05,
|
| 156 |
+
"loss": 1.2671,
|
| 157 |
+
"step": 760
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.4,
|
| 161 |
+
"learning_rate": 1.7355793772332825e-05,
|
| 162 |
+
"loss": 1.2515,
|
| 163 |
+
"step": 780
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.41,
|
| 167 |
+
"learning_rate": 1.7287731835970736e-05,
|
| 168 |
+
"loss": 1.2208,
|
| 169 |
+
"step": 800
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.42,
|
| 173 |
+
"learning_rate": 1.7219669899608644e-05,
|
| 174 |
+
"loss": 1.2493,
|
| 175 |
+
"step": 820
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.43,
|
| 179 |
+
"learning_rate": 1.7151607963246555e-05,
|
| 180 |
+
"loss": 1.2419,
|
| 181 |
+
"step": 840
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 0.44,
|
| 185 |
+
"learning_rate": 1.7083546026884466e-05,
|
| 186 |
+
"loss": 1.2693,
|
| 187 |
+
"step": 860
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 0.45,
|
| 191 |
+
"learning_rate": 1.7015484090522377e-05,
|
| 192 |
+
"loss": 1.2141,
|
| 193 |
+
"step": 880
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 0.46,
|
| 197 |
+
"learning_rate": 1.6947422154160288e-05,
|
| 198 |
+
"loss": 1.2465,
|
| 199 |
+
"step": 900
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.47,
|
| 203 |
+
"learning_rate": 1.68793602177982e-05,
|
| 204 |
+
"loss": 1.2586,
|
| 205 |
+
"step": 920
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.48,
|
| 209 |
+
"learning_rate": 1.681129828143611e-05,
|
| 210 |
+
"loss": 1.2394,
|
| 211 |
+
"step": 940
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 0.49,
|
| 215 |
+
"learning_rate": 1.6743236345074017e-05,
|
| 216 |
+
"loss": 1.2733,
|
| 217 |
+
"step": 960
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 0.5,
|
| 221 |
+
"learning_rate": 1.6675174408711928e-05,
|
| 222 |
+
"loss": 1.2734,
|
| 223 |
+
"step": 980
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 0.51,
|
| 227 |
+
"learning_rate": 1.660711247234984e-05,
|
| 228 |
+
"loss": 1.2333,
|
| 229 |
+
"step": 1000
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.52,
|
| 233 |
+
"learning_rate": 1.653905053598775e-05,
|
| 234 |
+
"loss": 1.198,
|
| 235 |
+
"step": 1020
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 0.53,
|
| 239 |
+
"learning_rate": 1.647098859962566e-05,
|
| 240 |
+
"loss": 1.2297,
|
| 241 |
+
"step": 1040
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.54,
|
| 245 |
+
"learning_rate": 1.6402926663263572e-05,
|
| 246 |
+
"loss": 1.2702,
|
| 247 |
+
"step": 1060
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.55,
|
| 251 |
+
"learning_rate": 1.6334864726901483e-05,
|
| 252 |
+
"loss": 1.2799,
|
| 253 |
+
"step": 1080
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 0.56,
|
| 257 |
+
"learning_rate": 1.6266802790539394e-05,
|
| 258 |
+
"loss": 1.2275,
|
| 259 |
+
"step": 1100
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 0.57,
|
| 263 |
+
"learning_rate": 1.61987408541773e-05,
|
| 264 |
+
"loss": 1.2433,
|
| 265 |
+
"step": 1120
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.58,
|
| 269 |
+
"learning_rate": 1.6130678917815213e-05,
|
| 270 |
+
"loss": 1.2376,
|
| 271 |
+
"step": 1140
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 0.59,
|
| 275 |
+
"learning_rate": 1.6062616981453123e-05,
|
| 276 |
+
"loss": 1.2032,
|
| 277 |
+
"step": 1160
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 0.6,
|
| 281 |
+
"learning_rate": 1.5994555045091034e-05,
|
| 282 |
+
"loss": 1.2623,
|
| 283 |
+
"step": 1180
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.61,
|
| 287 |
+
"learning_rate": 1.5926493108728945e-05,
|
| 288 |
+
"loss": 1.2367,
|
| 289 |
+
"step": 1200
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.62,
|
| 293 |
+
"learning_rate": 1.5858431172366856e-05,
|
| 294 |
+
"loss": 1.2015,
|
| 295 |
+
"step": 1220
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 0.63,
|
| 299 |
+
"learning_rate": 1.5790369236004764e-05,
|
| 300 |
+
"loss": 1.1896,
|
| 301 |
+
"step": 1240
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 0.64,
|
| 305 |
+
"learning_rate": 1.5722307299642675e-05,
|
| 306 |
+
"loss": 1.2868,
|
| 307 |
+
"step": 1260
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 0.65,
|
| 311 |
+
"learning_rate": 1.5654245363280586e-05,
|
| 312 |
+
"loss": 1.1849,
|
| 313 |
+
"step": 1280
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.66,
|
| 317 |
+
"learning_rate": 1.5586183426918497e-05,
|
| 318 |
+
"loss": 1.2463,
|
| 319 |
+
"step": 1300
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 0.67,
|
| 323 |
+
"learning_rate": 1.5518121490556408e-05,
|
| 324 |
+
"loss": 1.2507,
|
| 325 |
+
"step": 1320
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.68,
|
| 329 |
+
"learning_rate": 1.545005955419432e-05,
|
| 330 |
+
"loss": 1.2211,
|
| 331 |
+
"step": 1340
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.69,
|
| 335 |
+
"learning_rate": 1.538199761783223e-05,
|
| 336 |
+
"loss": 1.2082,
|
| 337 |
+
"step": 1360
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 0.7,
|
| 341 |
+
"learning_rate": 1.531393568147014e-05,
|
| 342 |
+
"loss": 1.2116,
|
| 343 |
+
"step": 1380
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 0.71,
|
| 347 |
+
"learning_rate": 1.5245873745108048e-05,
|
| 348 |
+
"loss": 1.2223,
|
| 349 |
+
"step": 1400
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.72,
|
| 353 |
+
"learning_rate": 1.517781180874596e-05,
|
| 354 |
+
"loss": 1.2209,
|
| 355 |
+
"step": 1420
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 0.74,
|
| 359 |
+
"learning_rate": 1.510974987238387e-05,
|
| 360 |
+
"loss": 1.2513,
|
| 361 |
+
"step": 1440
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.75,
|
| 365 |
+
"learning_rate": 1.5041687936021781e-05,
|
| 366 |
+
"loss": 1.1825,
|
| 367 |
+
"step": 1460
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.76,
|
| 371 |
+
"learning_rate": 1.4977029096477795e-05,
|
| 372 |
+
"loss": 1.2369,
|
| 373 |
+
"step": 1480
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.77,
|
| 377 |
+
"learning_rate": 1.4908967160115706e-05,
|
| 378 |
+
"loss": 1.2623,
|
| 379 |
+
"step": 1500
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 0.77,
|
| 383 |
+
"eval_gen_len": 10.2668,
|
| 384 |
+
"eval_loss": 1.3323159217834473,
|
| 385 |
+
"eval_rouge1": 5.5,
|
| 386 |
+
"eval_rouge2": 3.2122,
|
| 387 |
+
"eval_rougeL": 5.0222,
|
| 388 |
+
"eval_rougeLsum": 5.0534,
|
| 389 |
+
"eval_runtime": 141.5634,
|
| 390 |
+
"eval_samples_per_second": 35.56,
|
| 391 |
+
"eval_steps_per_second": 1.483,
|
| 392 |
+
"step": 1500
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.78,
|
| 396 |
+
"learning_rate": 1.4840905223753617e-05,
|
| 397 |
+
"loss": 1.2388,
|
| 398 |
+
"step": 1520
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.79,
|
| 402 |
+
"learning_rate": 1.4772843287391528e-05,
|
| 403 |
+
"loss": 1.2583,
|
| 404 |
+
"step": 1540
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.8,
|
| 408 |
+
"learning_rate": 1.4704781351029439e-05,
|
| 409 |
+
"loss": 1.1844,
|
| 410 |
+
"step": 1560
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.81,
|
| 414 |
+
"learning_rate": 1.4636719414667348e-05,
|
| 415 |
+
"loss": 1.2002,
|
| 416 |
+
"step": 1580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.82,
|
| 420 |
+
"learning_rate": 1.4568657478305259e-05,
|
| 421 |
+
"loss": 1.2349,
|
| 422 |
+
"step": 1600
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.83,
|
| 426 |
+
"learning_rate": 1.450059554194317e-05,
|
| 427 |
+
"loss": 1.2212,
|
| 428 |
+
"step": 1620
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.84,
|
| 432 |
+
"learning_rate": 1.443253360558108e-05,
|
| 433 |
+
"loss": 1.2169,
|
| 434 |
+
"step": 1640
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.85,
|
| 438 |
+
"learning_rate": 1.4364471669218992e-05,
|
| 439 |
+
"loss": 1.177,
|
| 440 |
+
"step": 1660
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.86,
|
| 444 |
+
"learning_rate": 1.4296409732856901e-05,
|
| 445 |
+
"loss": 1.1969,
|
| 446 |
+
"step": 1680
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.87,
|
| 450 |
+
"learning_rate": 1.4228347796494812e-05,
|
| 451 |
+
"loss": 1.2036,
|
| 452 |
+
"step": 1700
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.88,
|
| 456 |
+
"learning_rate": 1.4160285860132721e-05,
|
| 457 |
+
"loss": 1.1921,
|
| 458 |
+
"step": 1720
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.89,
|
| 462 |
+
"learning_rate": 1.4092223923770632e-05,
|
| 463 |
+
"loss": 1.2075,
|
| 464 |
+
"step": 1740
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.9,
|
| 468 |
+
"learning_rate": 1.4024161987408542e-05,
|
| 469 |
+
"loss": 1.2248,
|
| 470 |
+
"step": 1760
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.91,
|
| 474 |
+
"learning_rate": 1.3956100051046453e-05,
|
| 475 |
+
"loss": 1.1921,
|
| 476 |
+
"step": 1780
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.92,
|
| 480 |
+
"learning_rate": 1.3888038114684363e-05,
|
| 481 |
+
"loss": 1.2332,
|
| 482 |
+
"step": 1800
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.93,
|
| 486 |
+
"learning_rate": 1.3819976178322274e-05,
|
| 487 |
+
"loss": 1.2762,
|
| 488 |
+
"step": 1820
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.94,
|
| 492 |
+
"learning_rate": 1.3751914241960185e-05,
|
| 493 |
+
"loss": 1.24,
|
| 494 |
+
"step": 1840
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.95,
|
| 498 |
+
"learning_rate": 1.3683852305598095e-05,
|
| 499 |
+
"loss": 1.2117,
|
| 500 |
+
"step": 1860
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.96,
|
| 504 |
+
"learning_rate": 1.3615790369236006e-05,
|
| 505 |
+
"loss": 1.2361,
|
| 506 |
+
"step": 1880
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.97,
|
| 510 |
+
"learning_rate": 1.3547728432873917e-05,
|
| 511 |
+
"loss": 1.2175,
|
| 512 |
+
"step": 1900
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.98,
|
| 516 |
+
"learning_rate": 1.3479666496511828e-05,
|
| 517 |
+
"loss": 1.1877,
|
| 518 |
+
"step": 1920
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.99,
|
| 522 |
+
"learning_rate": 1.3411604560149739e-05,
|
| 523 |
+
"loss": 1.2637,
|
| 524 |
+
"step": 1940
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 1.0,
|
| 528 |
+
"learning_rate": 1.3343542623787648e-05,
|
| 529 |
+
"loss": 1.1752,
|
| 530 |
+
"step": 1960
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 1.01,
|
| 534 |
+
"learning_rate": 1.3275480687425559e-05,
|
| 535 |
+
"loss": 1.0205,
|
| 536 |
+
"step": 1980
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.02,
|
| 540 |
+
"learning_rate": 1.320741875106347e-05,
|
| 541 |
+
"loss": 1.0218,
|
| 542 |
+
"step": 2000
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.03,
|
| 546 |
+
"learning_rate": 1.3139356814701379e-05,
|
| 547 |
+
"loss": 1.0558,
|
| 548 |
+
"step": 2020
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.04,
|
| 552 |
+
"learning_rate": 1.3071294878339288e-05,
|
| 553 |
+
"loss": 1.0654,
|
| 554 |
+
"step": 2040
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 1.05,
|
| 558 |
+
"learning_rate": 1.30032329419772e-05,
|
| 559 |
+
"loss": 1.0294,
|
| 560 |
+
"step": 2060
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.06,
|
| 564 |
+
"learning_rate": 1.293517100561511e-05,
|
| 565 |
+
"loss": 1.0307,
|
| 566 |
+
"step": 2080
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 1.07,
|
| 570 |
+
"learning_rate": 1.2867109069253021e-05,
|
| 571 |
+
"loss": 1.0403,
|
| 572 |
+
"step": 2100
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 1.08,
|
| 576 |
+
"learning_rate": 1.2799047132890932e-05,
|
| 577 |
+
"loss": 1.0457,
|
| 578 |
+
"step": 2120
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.09,
|
| 582 |
+
"learning_rate": 1.2730985196528841e-05,
|
| 583 |
+
"loss": 0.9853,
|
| 584 |
+
"step": 2140
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.1,
|
| 588 |
+
"learning_rate": 1.2662923260166752e-05,
|
| 589 |
+
"loss": 1.0287,
|
| 590 |
+
"step": 2160
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.11,
|
| 594 |
+
"learning_rate": 1.2594861323804663e-05,
|
| 595 |
+
"loss": 1.0173,
|
| 596 |
+
"step": 2180
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 1.12,
|
| 600 |
+
"learning_rate": 1.2526799387442574e-05,
|
| 601 |
+
"loss": 1.0501,
|
| 602 |
+
"step": 2200
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 1.13,
|
| 606 |
+
"learning_rate": 1.2458737451080485e-05,
|
| 607 |
+
"loss": 1.0665,
|
| 608 |
+
"step": 2220
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 1.14,
|
| 612 |
+
"learning_rate": 1.2390675514718396e-05,
|
| 613 |
+
"loss": 1.0629,
|
| 614 |
+
"step": 2240
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 1.15,
|
| 618 |
+
"learning_rate": 1.2322613578356306e-05,
|
| 619 |
+
"loss": 1.0737,
|
| 620 |
+
"step": 2260
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.16,
|
| 624 |
+
"learning_rate": 1.2254551641994217e-05,
|
| 625 |
+
"loss": 1.0557,
|
| 626 |
+
"step": 2280
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.17,
|
| 630 |
+
"learning_rate": 1.2186489705632126e-05,
|
| 631 |
+
"loss": 1.0174,
|
| 632 |
+
"step": 2300
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.18,
|
| 636 |
+
"learning_rate": 1.2118427769270035e-05,
|
| 637 |
+
"loss": 1.0301,
|
| 638 |
+
"step": 2320
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 1.19,
|
| 642 |
+
"learning_rate": 1.2050365832907946e-05,
|
| 643 |
+
"loss": 1.0604,
|
| 644 |
+
"step": 2340
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 1.2,
|
| 648 |
+
"learning_rate": 1.1982303896545857e-05,
|
| 649 |
+
"loss": 1.0277,
|
| 650 |
+
"step": 2360
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 1.21,
|
| 654 |
+
"learning_rate": 1.1914241960183768e-05,
|
| 655 |
+
"loss": 1.0805,
|
| 656 |
+
"step": 2380
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 1.23,
|
| 660 |
+
"learning_rate": 1.1846180023821679e-05,
|
| 661 |
+
"loss": 1.0641,
|
| 662 |
+
"step": 2400
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.24,
|
| 666 |
+
"learning_rate": 1.1778118087459588e-05,
|
| 667 |
+
"loss": 1.0864,
|
| 668 |
+
"step": 2420
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.25,
|
| 672 |
+
"learning_rate": 1.17100561510975e-05,
|
| 673 |
+
"loss": 1.0462,
|
| 674 |
+
"step": 2440
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.26,
|
| 678 |
+
"learning_rate": 1.164199421473541e-05,
|
| 679 |
+
"loss": 1.0208,
|
| 680 |
+
"step": 2460
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 1.27,
|
| 684 |
+
"learning_rate": 1.1573932278373321e-05,
|
| 685 |
+
"loss": 1.0617,
|
| 686 |
+
"step": 2480
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 1.28,
|
| 690 |
+
"learning_rate": 1.1505870342011232e-05,
|
| 691 |
+
"loss": 1.039,
|
| 692 |
+
"step": 2500
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 1.29,
|
| 696 |
+
"learning_rate": 1.1437808405649143e-05,
|
| 697 |
+
"loss": 1.046,
|
| 698 |
+
"step": 2520
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 1.3,
|
| 702 |
+
"learning_rate": 1.1369746469287052e-05,
|
| 703 |
+
"loss": 1.0336,
|
| 704 |
+
"step": 2540
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.31,
|
| 708 |
+
"learning_rate": 1.1301684532924963e-05,
|
| 709 |
+
"loss": 1.0181,
|
| 710 |
+
"step": 2560
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.32,
|
| 714 |
+
"learning_rate": 1.1233622596562874e-05,
|
| 715 |
+
"loss": 1.0551,
|
| 716 |
+
"step": 2580
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.33,
|
| 720 |
+
"learning_rate": 1.1165560660200782e-05,
|
| 721 |
+
"loss": 1.0149,
|
| 722 |
+
"step": 2600
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 1.34,
|
| 726 |
+
"learning_rate": 1.1097498723838693e-05,
|
| 727 |
+
"loss": 1.0448,
|
| 728 |
+
"step": 2620
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 1.35,
|
| 732 |
+
"learning_rate": 1.1029436787476604e-05,
|
| 733 |
+
"loss": 1.0477,
|
| 734 |
+
"step": 2640
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 1.36,
|
| 738 |
+
"learning_rate": 1.0961374851114515e-05,
|
| 739 |
+
"loss": 1.0601,
|
| 740 |
+
"step": 2660
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 1.37,
|
| 744 |
+
"learning_rate": 1.0893312914752426e-05,
|
| 745 |
+
"loss": 1.0558,
|
| 746 |
+
"step": 2680
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.38,
|
| 750 |
+
"learning_rate": 1.0825250978390337e-05,
|
| 751 |
+
"loss": 1.0861,
|
| 752 |
+
"step": 2700
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.39,
|
| 756 |
+
"learning_rate": 1.0757189042028246e-05,
|
| 757 |
+
"loss": 1.0411,
|
| 758 |
+
"step": 2720
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.4,
|
| 762 |
+
"learning_rate": 1.0689127105666157e-05,
|
| 763 |
+
"loss": 1.0597,
|
| 764 |
+
"step": 2740
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 1.41,
|
| 768 |
+
"learning_rate": 1.0621065169304068e-05,
|
| 769 |
+
"loss": 1.0235,
|
| 770 |
+
"step": 2760
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 1.42,
|
| 774 |
+
"learning_rate": 1.0553003232941979e-05,
|
| 775 |
+
"loss": 1.063,
|
| 776 |
+
"step": 2780
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 1.43,
|
| 780 |
+
"learning_rate": 1.048494129657989e-05,
|
| 781 |
+
"loss": 1.0622,
|
| 782 |
+
"step": 2800
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 1.44,
|
| 786 |
+
"learning_rate": 1.0416879360217799e-05,
|
| 787 |
+
"loss": 1.0778,
|
| 788 |
+
"step": 2820
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.45,
|
| 792 |
+
"learning_rate": 1.034881742385571e-05,
|
| 793 |
+
"loss": 1.0465,
|
| 794 |
+
"step": 2840
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.46,
|
| 798 |
+
"learning_rate": 1.0280755487493621e-05,
|
| 799 |
+
"loss": 1.0644,
|
| 800 |
+
"step": 2860
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.47,
|
| 804 |
+
"learning_rate": 1.021269355113153e-05,
|
| 805 |
+
"loss": 1.0282,
|
| 806 |
+
"step": 2880
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 1.48,
|
| 810 |
+
"learning_rate": 1.014463161476944e-05,
|
| 811 |
+
"loss": 1.0326,
|
| 812 |
+
"step": 2900
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 1.49,
|
| 816 |
+
"learning_rate": 1.007656967840735e-05,
|
| 817 |
+
"loss": 1.0591,
|
| 818 |
+
"step": 2920
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 1.5,
|
| 822 |
+
"learning_rate": 1.0008507742045262e-05,
|
| 823 |
+
"loss": 1.0269,
|
| 824 |
+
"step": 2940
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 1.51,
|
| 828 |
+
"learning_rate": 9.940445805683172e-06,
|
| 829 |
+
"loss": 1.0326,
|
| 830 |
+
"step": 2960
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.52,
|
| 834 |
+
"learning_rate": 9.872383869321083e-06,
|
| 835 |
+
"loss": 1.0935,
|
| 836 |
+
"step": 2980
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.53,
|
| 840 |
+
"learning_rate": 9.804321932958993e-06,
|
| 841 |
+
"loss": 1.0214,
|
| 842 |
+
"step": 3000
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.53,
|
| 846 |
+
"eval_gen_len": 12.1136,
|
| 847 |
+
"eval_loss": 1.3700170516967773,
|
| 848 |
+
"eval_rouge1": 6.4308,
|
| 849 |
+
"eval_rouge2": 3.9131,
|
| 850 |
+
"eval_rougeL": 6.0216,
|
| 851 |
+
"eval_rougeLsum": 6.0417,
|
| 852 |
+
"eval_runtime": 143.5844,
|
| 853 |
+
"eval_samples_per_second": 35.06,
|
| 854 |
+
"eval_steps_per_second": 1.463,
|
| 855 |
+
"step": 3000
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 1.54,
|
| 859 |
+
"learning_rate": 9.736259996596904e-06,
|
| 860 |
+
"loss": 1.0338,
|
| 861 |
+
"step": 3020
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 1.55,
|
| 865 |
+
"learning_rate": 9.668198060234815e-06,
|
| 866 |
+
"loss": 1.046,
|
| 867 |
+
"step": 3040
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 1.56,
|
| 871 |
+
"learning_rate": 9.600136123872726e-06,
|
| 872 |
+
"loss": 1.0208,
|
| 873 |
+
"step": 3060
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 1.57,
|
| 877 |
+
"learning_rate": 9.532074187510637e-06,
|
| 878 |
+
"loss": 1.0093,
|
| 879 |
+
"step": 3080
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 1.58,
|
| 883 |
+
"learning_rate": 9.464012251148546e-06,
|
| 884 |
+
"loss": 1.0358,
|
| 885 |
+
"step": 3100
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.59,
|
| 889 |
+
"learning_rate": 9.395950314786457e-06,
|
| 890 |
+
"loss": 1.0709,
|
| 891 |
+
"step": 3120
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.6,
|
| 895 |
+
"learning_rate": 9.327888378424366e-06,
|
| 896 |
+
"loss": 1.0602,
|
| 897 |
+
"step": 3140
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 1.61,
|
| 901 |
+
"learning_rate": 9.259826442062277e-06,
|
| 902 |
+
"loss": 1.0062,
|
| 903 |
+
"step": 3160
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 1.62,
|
| 907 |
+
"learning_rate": 9.191764505700188e-06,
|
| 908 |
+
"loss": 1.0823,
|
| 909 |
+
"step": 3180
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 1.63,
|
| 913 |
+
"learning_rate": 9.123702569338099e-06,
|
| 914 |
+
"loss": 1.0439,
|
| 915 |
+
"step": 3200
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 1.64,
|
| 919 |
+
"learning_rate": 9.05564063297601e-06,
|
| 920 |
+
"loss": 1.0454,
|
| 921 |
+
"step": 3220
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 1.65,
|
| 925 |
+
"learning_rate": 8.98757869661392e-06,
|
| 926 |
+
"loss": 1.0806,
|
| 927 |
+
"step": 3240
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.66,
|
| 931 |
+
"learning_rate": 8.91951676025183e-06,
|
| 932 |
+
"loss": 1.0176,
|
| 933 |
+
"step": 3260
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.67,
|
| 937 |
+
"learning_rate": 8.85145482388974e-06,
|
| 938 |
+
"loss": 1.0799,
|
| 939 |
+
"step": 3280
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 1.68,
|
| 943 |
+
"learning_rate": 8.78339288752765e-06,
|
| 944 |
+
"loss": 1.0421,
|
| 945 |
+
"step": 3300
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 1.69,
|
| 949 |
+
"learning_rate": 8.715330951165561e-06,
|
| 950 |
+
"loss": 1.0136,
|
| 951 |
+
"step": 3320
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 1.7,
|
| 955 |
+
"learning_rate": 8.647269014803472e-06,
|
| 956 |
+
"loss": 0.9866,
|
| 957 |
+
"step": 3340
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 1.72,
|
| 961 |
+
"learning_rate": 8.579207078441383e-06,
|
| 962 |
+
"loss": 1.0355,
|
| 963 |
+
"step": 3360
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 1.73,
|
| 967 |
+
"learning_rate": 8.511145142079293e-06,
|
| 968 |
+
"loss": 1.0791,
|
| 969 |
+
"step": 3380
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.74,
|
| 973 |
+
"learning_rate": 8.443083205717204e-06,
|
| 974 |
+
"loss": 1.0524,
|
| 975 |
+
"step": 3400
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.75,
|
| 979 |
+
"learning_rate": 8.375021269355113e-06,
|
| 980 |
+
"loss": 1.0669,
|
| 981 |
+
"step": 3420
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 1.76,
|
| 985 |
+
"learning_rate": 8.306959332993024e-06,
|
| 986 |
+
"loss": 1.007,
|
| 987 |
+
"step": 3440
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 1.77,
|
| 991 |
+
"learning_rate": 8.238897396630935e-06,
|
| 992 |
+
"loss": 1.0121,
|
| 993 |
+
"step": 3460
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 1.78,
|
| 997 |
+
"learning_rate": 8.170835460268846e-06,
|
| 998 |
+
"loss": 1.0281,
|
| 999 |
+
"step": 3480
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 1.79,
|
| 1003 |
+
"learning_rate": 8.102773523906757e-06,
|
| 1004 |
+
"loss": 1.062,
|
| 1005 |
+
"step": 3500
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 1.8,
|
| 1009 |
+
"learning_rate": 8.034711587544666e-06,
|
| 1010 |
+
"loss": 1.0041,
|
| 1011 |
+
"step": 3520
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.81,
|
| 1015 |
+
"learning_rate": 7.966649651182577e-06,
|
| 1016 |
+
"loss": 1.043,
|
| 1017 |
+
"step": 3540
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.82,
|
| 1021 |
+
"learning_rate": 7.898587714820486e-06,
|
| 1022 |
+
"loss": 1.0586,
|
| 1023 |
+
"step": 3560
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 1.83,
|
| 1027 |
+
"learning_rate": 7.830525778458397e-06,
|
| 1028 |
+
"loss": 1.0733,
|
| 1029 |
+
"step": 3580
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 1.84,
|
| 1033 |
+
"learning_rate": 7.762463842096308e-06,
|
| 1034 |
+
"loss": 1.0441,
|
| 1035 |
+
"step": 3600
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 1.85,
|
| 1039 |
+
"learning_rate": 7.69440190573422e-06,
|
| 1040 |
+
"loss": 1.0267,
|
| 1041 |
+
"step": 3620
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 1.86,
|
| 1045 |
+
"learning_rate": 7.626339969372129e-06,
|
| 1046 |
+
"loss": 1.0521,
|
| 1047 |
+
"step": 3640
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 1.87,
|
| 1051 |
+
"learning_rate": 7.55827803301004e-06,
|
| 1052 |
+
"loss": 1.0168,
|
| 1053 |
+
"step": 3660
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.88,
|
| 1057 |
+
"learning_rate": 7.4902160966479495e-06,
|
| 1058 |
+
"loss": 1.046,
|
| 1059 |
+
"step": 3680
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.89,
|
| 1063 |
+
"learning_rate": 7.4221541602858605e-06,
|
| 1064 |
+
"loss": 1.0259,
|
| 1065 |
+
"step": 3700
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 1.9,
|
| 1069 |
+
"learning_rate": 7.3540922239237715e-06,
|
| 1070 |
+
"loss": 1.0173,
|
| 1071 |
+
"step": 3720
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 1.91,
|
| 1075 |
+
"learning_rate": 7.286030287561682e-06,
|
| 1076 |
+
"loss": 1.0507,
|
| 1077 |
+
"step": 3740
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 1.92,
|
| 1081 |
+
"learning_rate": 7.2179683511995926e-06,
|
| 1082 |
+
"loss": 1.0145,
|
| 1083 |
+
"step": 3760
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 1.93,
|
| 1087 |
+
"learning_rate": 7.149906414837503e-06,
|
| 1088 |
+
"loss": 1.0193,
|
| 1089 |
+
"step": 3780
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 1.94,
|
| 1093 |
+
"learning_rate": 7.081844478475414e-06,
|
| 1094 |
+
"loss": 1.0452,
|
| 1095 |
+
"step": 3800
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.95,
|
| 1099 |
+
"learning_rate": 7.013782542113323e-06,
|
| 1100 |
+
"loss": 1.0494,
|
| 1101 |
+
"step": 3820
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.96,
|
| 1105 |
+
"learning_rate": 6.945720605751234e-06,
|
| 1106 |
+
"loss": 1.0334,
|
| 1107 |
+
"step": 3840
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 1.97,
|
| 1111 |
+
"learning_rate": 6.877658669389145e-06,
|
| 1112 |
+
"loss": 0.9954,
|
| 1113 |
+
"step": 3860
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 1.98,
|
| 1117 |
+
"learning_rate": 6.809596733027055e-06,
|
| 1118 |
+
"loss": 1.0286,
|
| 1119 |
+
"step": 3880
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 1.99,
|
| 1123 |
+
"learning_rate": 6.741534796664966e-06,
|
| 1124 |
+
"loss": 1.0547,
|
| 1125 |
+
"step": 3900
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 2.0,
|
| 1129 |
+
"learning_rate": 6.673472860302876e-06,
|
| 1130 |
+
"loss": 1.0648,
|
| 1131 |
+
"step": 3920
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 2.01,
|
| 1135 |
+
"learning_rate": 6.605410923940787e-06,
|
| 1136 |
+
"loss": 0.9251,
|
| 1137 |
+
"step": 3940
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 2.02,
|
| 1141 |
+
"learning_rate": 6.537348987578696e-06,
|
| 1142 |
+
"loss": 0.9152,
|
| 1143 |
+
"step": 3960
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 2.03,
|
| 1147 |
+
"learning_rate": 6.469287051216607e-06,
|
| 1148 |
+
"loss": 0.9066,
|
| 1149 |
+
"step": 3980
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 2.04,
|
| 1153 |
+
"learning_rate": 6.401225114854518e-06,
|
| 1154 |
+
"loss": 0.8861,
|
| 1155 |
+
"step": 4000
|
| 1156 |
+
},
|
| 1157 |
+
{
|
| 1158 |
+
"epoch": 2.05,
|
| 1159 |
+
"learning_rate": 6.333163178492428e-06,
|
| 1160 |
+
"loss": 0.8827,
|
| 1161 |
+
"step": 4020
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 2.06,
|
| 1165 |
+
"learning_rate": 6.265101242130339e-06,
|
| 1166 |
+
"loss": 0.9346,
|
| 1167 |
+
"step": 4040
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 2.07,
|
| 1171 |
+
"learning_rate": 6.1970393057682494e-06,
|
| 1172 |
+
"loss": 0.9073,
|
| 1173 |
+
"step": 4060
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 2.08,
|
| 1177 |
+
"learning_rate": 6.12897736940616e-06,
|
| 1178 |
+
"loss": 0.9338,
|
| 1179 |
+
"step": 4080
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 2.09,
|
| 1183 |
+
"learning_rate": 6.060915433044071e-06,
|
| 1184 |
+
"loss": 0.9163,
|
| 1185 |
+
"step": 4100
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 2.1,
|
| 1189 |
+
"learning_rate": 5.992853496681981e-06,
|
| 1190 |
+
"loss": 0.9198,
|
| 1191 |
+
"step": 4120
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 2.11,
|
| 1195 |
+
"learning_rate": 5.924791560319892e-06,
|
| 1196 |
+
"loss": 0.9399,
|
| 1197 |
+
"step": 4140
|
| 1198 |
+
},
|
| 1199 |
+
{
|
| 1200 |
+
"epoch": 2.12,
|
| 1201 |
+
"learning_rate": 5.856729623957802e-06,
|
| 1202 |
+
"loss": 0.8875,
|
| 1203 |
+
"step": 4160
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 2.13,
|
| 1207 |
+
"learning_rate": 5.788667687595713e-06,
|
| 1208 |
+
"loss": 0.926,
|
| 1209 |
+
"step": 4180
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 2.14,
|
| 1213 |
+
"learning_rate": 5.720605751233624e-06,
|
| 1214 |
+
"loss": 0.9618,
|
| 1215 |
+
"step": 4200
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 2.15,
|
| 1219 |
+
"learning_rate": 5.652543814871534e-06,
|
| 1220 |
+
"loss": 0.9256,
|
| 1221 |
+
"step": 4220
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 2.16,
|
| 1225 |
+
"learning_rate": 5.587884975327549e-06,
|
| 1226 |
+
"loss": 0.9011,
|
| 1227 |
+
"step": 4240
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 2.17,
|
| 1231 |
+
"learning_rate": 5.519823038965459e-06,
|
| 1232 |
+
"loss": 0.9066,
|
| 1233 |
+
"step": 4260
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 2.18,
|
| 1237 |
+
"learning_rate": 5.451761102603369e-06,
|
| 1238 |
+
"loss": 0.8843,
|
| 1239 |
+
"step": 4280
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 2.19,
|
| 1243 |
+
"learning_rate": 5.383699166241279e-06,
|
| 1244 |
+
"loss": 0.9288,
|
| 1245 |
+
"step": 4300
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 2.21,
|
| 1249 |
+
"learning_rate": 5.31563722987919e-06,
|
| 1250 |
+
"loss": 0.9352,
|
| 1251 |
+
"step": 4320
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 2.22,
|
| 1255 |
+
"learning_rate": 5.247575293517101e-06,
|
| 1256 |
+
"loss": 0.937,
|
| 1257 |
+
"step": 4340
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 2.23,
|
| 1261 |
+
"learning_rate": 5.1795133571550115e-06,
|
| 1262 |
+
"loss": 0.9036,
|
| 1263 |
+
"step": 4360
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 2.24,
|
| 1267 |
+
"learning_rate": 5.1114514207929224e-06,
|
| 1268 |
+
"loss": 0.8833,
|
| 1269 |
+
"step": 4380
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 2.25,
|
| 1273 |
+
"learning_rate": 5.0433894844308325e-06,
|
| 1274 |
+
"loss": 0.9567,
|
| 1275 |
+
"step": 4400
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 2.26,
|
| 1279 |
+
"learning_rate": 4.975327548068743e-06,
|
| 1280 |
+
"loss": 0.9222,
|
| 1281 |
+
"step": 4420
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 2.27,
|
| 1285 |
+
"learning_rate": 4.907265611706654e-06,
|
| 1286 |
+
"loss": 0.9528,
|
| 1287 |
+
"step": 4440
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 2.28,
|
| 1291 |
+
"learning_rate": 4.839203675344564e-06,
|
| 1292 |
+
"loss": 0.9287,
|
| 1293 |
+
"step": 4460
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 2.29,
|
| 1297 |
+
"learning_rate": 4.771141738982475e-06,
|
| 1298 |
+
"loss": 0.9269,
|
| 1299 |
+
"step": 4480
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 2.3,
|
| 1303 |
+
"learning_rate": 4.703079802620385e-06,
|
| 1304 |
+
"loss": 0.8977,
|
| 1305 |
+
"step": 4500
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 2.3,
|
| 1309 |
+
"eval_gen_len": 12.9261,
|
| 1310 |
+
"eval_loss": 1.3723982572555542,
|
| 1311 |
+
"eval_rouge1": 4.2774,
|
| 1312 |
+
"eval_rouge2": 2.4929,
|
| 1313 |
+
"eval_rougeL": 3.9376,
|
| 1314 |
+
"eval_rougeLsum": 3.9429,
|
| 1315 |
+
"eval_runtime": 142.026,
|
| 1316 |
+
"eval_samples_per_second": 35.444,
|
| 1317 |
+
"eval_steps_per_second": 1.479,
|
| 1318 |
+
"step": 4500
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 2.31,
|
| 1322 |
+
"learning_rate": 4.635017866258296e-06,
|
| 1323 |
+
"loss": 0.913,
|
| 1324 |
+
"step": 4520
|
| 1325 |
+
},
|
| 1326 |
+
{
|
| 1327 |
+
"epoch": 2.32,
|
| 1328 |
+
"learning_rate": 4.566955929896206e-06,
|
| 1329 |
+
"loss": 0.9012,
|
| 1330 |
+
"step": 4540
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 2.33,
|
| 1334 |
+
"learning_rate": 4.498893993534116e-06,
|
| 1335 |
+
"loss": 0.9077,
|
| 1336 |
+
"step": 4560
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 2.34,
|
| 1340 |
+
"learning_rate": 4.430832057172027e-06,
|
| 1341 |
+
"loss": 0.9181,
|
| 1342 |
+
"step": 4580
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 2.35,
|
| 1346 |
+
"learning_rate": 4.362770120809938e-06,
|
| 1347 |
+
"loss": 0.9234,
|
| 1348 |
+
"step": 4600
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 2.36,
|
| 1352 |
+
"learning_rate": 4.294708184447848e-06,
|
| 1353 |
+
"loss": 0.9014,
|
| 1354 |
+
"step": 4620
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 2.37,
|
| 1358 |
+
"learning_rate": 4.226646248085758e-06,
|
| 1359 |
+
"loss": 0.8963,
|
| 1360 |
+
"step": 4640
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 2.38,
|
| 1364 |
+
"learning_rate": 4.158584311723669e-06,
|
| 1365 |
+
"loss": 0.9097,
|
| 1366 |
+
"step": 4660
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 2.39,
|
| 1370 |
+
"learning_rate": 4.090522375361579e-06,
|
| 1371 |
+
"loss": 0.903,
|
| 1372 |
+
"step": 4680
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 2.4,
|
| 1376 |
+
"learning_rate": 4.0224604389994894e-06,
|
| 1377 |
+
"loss": 0.8938,
|
| 1378 |
+
"step": 4700
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 2.41,
|
| 1382 |
+
"learning_rate": 3.9543985026374e-06,
|
| 1383 |
+
"loss": 0.9029,
|
| 1384 |
+
"step": 4720
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 2.42,
|
| 1388 |
+
"learning_rate": 3.886336566275311e-06,
|
| 1389 |
+
"loss": 0.9338,
|
| 1390 |
+
"step": 4740
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 2.43,
|
| 1394 |
+
"learning_rate": 3.8182746299132215e-06,
|
| 1395 |
+
"loss": 0.9295,
|
| 1396 |
+
"step": 4760
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 2.44,
|
| 1400 |
+
"learning_rate": 3.7502126935511316e-06,
|
| 1401 |
+
"loss": 0.8946,
|
| 1402 |
+
"step": 4780
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 2.45,
|
| 1406 |
+
"learning_rate": 3.6821507571890426e-06,
|
| 1407 |
+
"loss": 0.891,
|
| 1408 |
+
"step": 4800
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 2.46,
|
| 1412 |
+
"learning_rate": 3.6140888208269527e-06,
|
| 1413 |
+
"loss": 0.9123,
|
| 1414 |
+
"step": 4820
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 2.47,
|
| 1418 |
+
"learning_rate": 3.5460268844648632e-06,
|
| 1419 |
+
"loss": 0.9518,
|
| 1420 |
+
"step": 4840
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 2.48,
|
| 1424 |
+
"learning_rate": 3.477964948102774e-06,
|
| 1425 |
+
"loss": 0.9378,
|
| 1426 |
+
"step": 4860
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 2.49,
|
| 1430 |
+
"learning_rate": 3.4099030117406843e-06,
|
| 1431 |
+
"loss": 0.9089,
|
| 1432 |
+
"step": 4880
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 2.5,
|
| 1436 |
+
"learning_rate": 3.3418410753785945e-06,
|
| 1437 |
+
"loss": 0.9241,
|
| 1438 |
+
"step": 4900
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 2.51,
|
| 1442 |
+
"learning_rate": 3.2737791390165054e-06,
|
| 1443 |
+
"loss": 0.9281,
|
| 1444 |
+
"step": 4920
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 2.52,
|
| 1448 |
+
"learning_rate": 3.205717202654416e-06,
|
| 1449 |
+
"loss": 0.9082,
|
| 1450 |
+
"step": 4940
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 2.53,
|
| 1454 |
+
"learning_rate": 3.1376552662923265e-06,
|
| 1455 |
+
"loss": 0.915,
|
| 1456 |
+
"step": 4960
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 2.54,
|
| 1460 |
+
"learning_rate": 3.0695933299302366e-06,
|
| 1461 |
+
"loss": 0.882,
|
| 1462 |
+
"step": 4980
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 2.55,
|
| 1466 |
+
"learning_rate": 3.001531393568147e-06,
|
| 1467 |
+
"loss": 0.9182,
|
| 1468 |
+
"step": 5000
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 2.56,
|
| 1472 |
+
"learning_rate": 2.9334694572060577e-06,
|
| 1473 |
+
"loss": 0.9369,
|
| 1474 |
+
"step": 5020
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 2.57,
|
| 1478 |
+
"learning_rate": 2.865407520843968e-06,
|
| 1479 |
+
"loss": 0.9324,
|
| 1480 |
+
"step": 5040
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 2.58,
|
| 1484 |
+
"learning_rate": 2.797345584481879e-06,
|
| 1485 |
+
"loss": 0.9069,
|
| 1486 |
+
"step": 5060
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 2.59,
|
| 1490 |
+
"learning_rate": 2.7292836481197893e-06,
|
| 1491 |
+
"loss": 0.9079,
|
| 1492 |
+
"step": 5080
|
| 1493 |
+
},
|
| 1494 |
+
{
|
| 1495 |
+
"epoch": 2.6,
|
| 1496 |
+
"learning_rate": 2.6612217117577e-06,
|
| 1497 |
+
"loss": 0.8965,
|
| 1498 |
+
"step": 5100
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 2.61,
|
| 1502 |
+
"learning_rate": 2.59315977539561e-06,
|
| 1503 |
+
"loss": 0.9312,
|
| 1504 |
+
"step": 5120
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 2.62,
|
| 1508 |
+
"learning_rate": 2.5250978390335206e-06,
|
| 1509 |
+
"loss": 0.9454,
|
| 1510 |
+
"step": 5140
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 2.63,
|
| 1514 |
+
"learning_rate": 2.4570359026714315e-06,
|
| 1515 |
+
"loss": 0.9115,
|
| 1516 |
+
"step": 5160
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 2.64,
|
| 1520 |
+
"learning_rate": 2.3889739663093416e-06,
|
| 1521 |
+
"loss": 0.9265,
|
| 1522 |
+
"step": 5180
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 2.65,
|
| 1526 |
+
"learning_rate": 2.320912029947252e-06,
|
| 1527 |
+
"loss": 0.9381,
|
| 1528 |
+
"step": 5200
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 2.66,
|
| 1532 |
+
"learning_rate": 2.2528500935851627e-06,
|
| 1533 |
+
"loss": 0.9334,
|
| 1534 |
+
"step": 5220
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 2.67,
|
| 1538 |
+
"learning_rate": 2.1847881572230733e-06,
|
| 1539 |
+
"loss": 0.914,
|
| 1540 |
+
"step": 5240
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 2.69,
|
| 1544 |
+
"learning_rate": 2.116726220860984e-06,
|
| 1545 |
+
"loss": 0.8805,
|
| 1546 |
+
"step": 5260
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 2.7,
|
| 1550 |
+
"learning_rate": 2.048664284498894e-06,
|
| 1551 |
+
"loss": 0.9417,
|
| 1552 |
+
"step": 5280
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 2.71,
|
| 1556 |
+
"learning_rate": 1.980602348136805e-06,
|
| 1557 |
+
"loss": 0.8974,
|
| 1558 |
+
"step": 5300
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 2.72,
|
| 1562 |
+
"learning_rate": 1.912540411774715e-06,
|
| 1563 |
+
"loss": 0.9722,
|
| 1564 |
+
"step": 5320
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 2.73,
|
| 1568 |
+
"learning_rate": 1.8444784754126258e-06,
|
| 1569 |
+
"loss": 0.9155,
|
| 1570 |
+
"step": 5340
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 2.74,
|
| 1574 |
+
"learning_rate": 1.7764165390505361e-06,
|
| 1575 |
+
"loss": 0.9088,
|
| 1576 |
+
"step": 5360
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"epoch": 2.75,
|
| 1580 |
+
"learning_rate": 1.7083546026884464e-06,
|
| 1581 |
+
"loss": 0.894,
|
| 1582 |
+
"step": 5380
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"epoch": 2.76,
|
| 1586 |
+
"learning_rate": 1.6402926663263572e-06,
|
| 1587 |
+
"loss": 0.936,
|
| 1588 |
+
"step": 5400
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 2.77,
|
| 1592 |
+
"learning_rate": 1.5722307299642675e-06,
|
| 1593 |
+
"loss": 0.8867,
|
| 1594 |
+
"step": 5420
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 2.78,
|
| 1598 |
+
"learning_rate": 1.504168793602178e-06,
|
| 1599 |
+
"loss": 0.9142,
|
| 1600 |
+
"step": 5440
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 2.79,
|
| 1604 |
+
"learning_rate": 1.4361068572400886e-06,
|
| 1605 |
+
"loss": 0.9394,
|
| 1606 |
+
"step": 5460
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 2.8,
|
| 1610 |
+
"learning_rate": 1.3680449208779992e-06,
|
| 1611 |
+
"loss": 0.9313,
|
| 1612 |
+
"step": 5480
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 2.81,
|
| 1616 |
+
"learning_rate": 1.2999829845159095e-06,
|
| 1617 |
+
"loss": 0.9197,
|
| 1618 |
+
"step": 5500
|
| 1619 |
+
}
|
| 1620 |
+
],
|
| 1621 |
+
"max_steps": 5877,
|
| 1622 |
+
"num_train_epochs": 3,
|
| 1623 |
+
"total_flos": 1.5087907085549568e+16,
|
| 1624 |
+
"trial_name": null,
|
| 1625 |
+
"trial_params": null
|
| 1626 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:937470ae204bd04d9ec5b79d9cac84bc4bb3aa6426fd0e9fcf85b202810a72e3
|
| 3 |
+
size 3119
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|